effect of substituents on the excited-state dynamics of …devries/groupsite/pub/24dapy...

14
Effect of substituents on the excited-state dynamics of the modified DNA bases 2,4-diaminopyrimidine and 2,6-diaminopurinew Zsolt Gengeliczki, a Michael P. Callahan, c Nathan Svadlenak, c Csaba Istva´n Pongor, b Ba´lint Szta´ray, bd Leo Meerts, g Dana Nachtigallova´,* e Pavel Hobza, e Mario Barbatti, f Hans Lischka* ef and Mattanjah S. de Vries* c Received 1st September 2009, Accepted 9th February 2010 First published as an Advance Article on the web 8th April 2010 DOI: 10.1039/b917852j To explore the excited state dynamics of pyrimidine derivatives, we performed a combined experimental and theoretical study. We present resonant two-photon ionization (R2PI) and IR-UV double resonance spectra of 2,4-diaminopyrimidine and 2,6-diaminopurine seeded in a supersonic jet by laser desorption. For 2,4-diaminopyrimidine (S 0 - S 1 34 459 cm 1 ), we observed only the diamino tautomer with an excited state lifetime bracketed between experimental limits of 10 ps and 1 ns. For 2,6-diaminopurine, we observed two tautomers, the 9H- (S 0 - S 1 34 881 cm 1 ) and 7H- (S 0 - S 1 32 215 cm 1 ) diamino forms, with excited state lifetimes of 6.3 0.4 ns and 8.7 0.8 ns, respectively. We investigated the nature of the excited state of 2,4-diaminopyrimidine by means of multi-reference ab initio methods. The calculations of stationary points in the ground and excited states, minima on the S 0 /S 1 crossing seam and connecting reaction paths show that several paths with negligible barriers exist, allowing ultrafast radiationless deactivation if excited at energies slightly higher than the band origin. The sub-nanosecond lifetime found experimentally is in good agreement with this finding. Introduction Gas phase laser spectroscopy provides the means to study the intrinsic properties of biologically relevant molecules in iso- lation. Such studies on RNA and DNA bases have revealed unique photophysical properties that are sensitive to subtle structural differences. In many cases, the biologically most relevant tautomeric form has a sub picosecond excited state lifetime, while other tautomeric forms of the same compound are much longer lived. 1,2 We have even found that the Watson–Crick structure, adopted by the guanine-cytosine (GC) base pair in DNA, appears to have a much shorter excited state lifetime than other structures of the same base pair. 3 Theoretical models explain these short lifetimes by a rapid internal conversion in which the excited state (S 1 ) is coupled to the ground state (S 0 ) via pathways with no or a very small barrier leading to conical intersections. 4 For the longer lived structures, small differences in relative energies cause the existence of barriers that lead to discrete spectra and lifetimes that can be two orders of magnitude longer. This rapid internal conversion pathway provides selected isomers with significantly enhanced photochemical stability, absent in the other longer lived structures. It is conceivable that these differences between excited state lifetimes of different bases and base-pair structures could have played a significant role in prebiotic chemistry. 3 The five naturally occurring nucleic acid bases 5–9 exhibit an ultrafast excited state relaxation in the gas phase. For DNA bases it has been shown 1,10–24 that the energetically lowest conical intersections responsible for fast deactivation to the ground state are characterized by ring puckering modes. The most favorable ring puckering conical intersections result from the change of the HC6C5R (R = H, CH 3 ) dihedral angle in uracil and thymine and the HN1C2H dihedral angle in adenine. This model also explains why by contrast 2-aminopurine, in which the C2 position is substituted by the amino group, has a long excited state lifetime and strong fluorescence. 25–27 Alternatively, the puckering at the C6 atom was also suggested to be responsible for adenine relaxation. 28 Recently, surface-hopping dynamics studies of 4-amino- pyrimidine (4-APy) 29–31 and 9H-adenine 24 showed that while the former relaxes into the ground state via different conical a Department of Chemistry, Stanford University, Stanford, CA 94305-5080, USA. E-mail: [email protected], [email protected], [email protected]; Fax: (+1) 805-893-4120; Tel: (+1) 805-893-5921 b Institute of Chemistry, Eo ¨tvo ¨s Lora ´nd University Budapest, 1/A Pa ´zma ´ny P. stny., Hungary 1117 c Department of Chemistry and Biochemistry, University of California, Santa Barbara, CA-93106-9510, USA d Department of Chemistry, University of the Pacific, Stockton, CA-95211, USA e The Institute of Organic Chemistry and Biochemistry, Flemingovo na ´m., 2, 166 10 Praha 6, Czech Republic f Institute for Theoretical Chemistry, University of Vienna, Waehringerstrasse 17, A1090 Vienna, Austria g Molecular and Biophysics Group, Institute for Molecules and Materials, Radboud University, 6500 GL Nijmegen, The Netherlands w Electronic supplementary information (ESI) available: The calcu- lated vertical ionization potentials for all the 2,4-diaminopyrimidine tautomers (Fig. S1 and Table S1). The vertical excitation energies calculated at the RICC2 method with various basis sets (Table S2). The interpolation curves between the S1min_C2 and S1min_C6 minima towards various MXS structures calculated at the CASSCF and MR-CISD(17)+Q methods (Fig. S2). This material is available free of charge on the Internet. See DOI: 10.1039/b917852j This journal is c the Owner Societies 2010 Phys. Chem. Chem. Phys., 2010, 12, 5375–5388 | 5375 PAPER www.rsc.org/pccp | Physical Chemistry Chemical Physics

Upload: vanthuy

Post on 01-Sep-2018

214 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Effect of substituents on the excited-state dynamics of …devries/groupsite/pub/24DAPy 26DAPu.pdf · previously studied the nucleobase xanthine.46 In this paper we ... pseudo-Voigt

Effect of substituents on the excited-state dynamics of the modified DNA

bases 2,4-diaminopyrimidine and 2,6-diaminopurinew

Zsolt Gengeliczki,a Michael P. Callahan,c Nathan Svadlenak,c

Csaba Istvan Pongor,bBalint Sztaray,

bdLeo Meerts,

gDana Nachtigallova,*

e

Pavel Hobza,eMario Barbatti,

fHans Lischka*

efand Mattanjah S. de Vries*

c

Received 1st September 2009, Accepted 9th February 2010

First published as an Advance Article on the web 8th April 2010

DOI: 10.1039/b917852j

To explore the excited state dynamics of pyrimidine derivatives, we performed a combined

experimental and theoretical study. We present resonant two-photon ionization (R2PI) and

IR-UV double resonance spectra of 2,4-diaminopyrimidine and 2,6-diaminopurine seeded

in a supersonic jet by laser desorption. For 2,4-diaminopyrimidine (S0 - S1 34 459 cm�1),

we observed only the diamino tautomer with an excited state lifetime bracketed between

experimental limits of 10 ps and 1 ns. For 2,6-diaminopurine, we observed two tautomers, the

9H- (S0 - S1 34 881 cm�1) and 7H- (S0 - S1 32 215 cm�1) diamino forms, with excited state

lifetimes of 6.3 � 0.4 ns and 8.7 � 0.8 ns, respectively. We investigated the nature of the excited

state of 2,4-diaminopyrimidine by means of multi-reference ab initio methods. The calculations

of stationary points in the ground and excited states, minima on the S0/S1 crossing seam and

connecting reaction paths show that several paths with negligible barriers exist, allowing

ultrafast radiationless deactivation if excited at energies slightly higher than the band origin.

The sub-nanosecond lifetime found experimentally is in good agreement with this finding.

Introduction

Gas phase laser spectroscopy provides the means to study the

intrinsic properties of biologically relevant molecules in iso-

lation. Such studies on RNA and DNA bases have revealed

unique photophysical properties that are sensitive to subtle

structural differences. In many cases, the biologically most

relevant tautomeric form has a sub picosecond excited state

lifetime, while other tautomeric forms of the same compound

are much longer lived.1,2 We have even found that the

Watson–Crick structure, adopted by the guanine-cytosine

(GC) base pair in DNA, appears to have a much shorter

excited state lifetime than other structures of the same base

pair.3 Theoretical models explain these short lifetimes by a

rapid internal conversion in which the excited state (S1) is

coupled to the ground state (S0) via pathways with no or a very

small barrier leading to conical intersections.4 For the longer

lived structures, small differences in relative energies cause the

existence of barriers that lead to discrete spectra and lifetimes

that can be two orders of magnitude longer. This rapid

internal conversion pathway provides selected isomers with

significantly enhanced photochemical stability, absent in the

other longer lived structures. It is conceivable that these

differences between excited state lifetimes of different bases

and base-pair structures could have played a significant role in

prebiotic chemistry.3

The five naturally occurring nucleic acid bases5–9 exhibit an

ultrafast excited state relaxation in the gas phase. For DNA

bases it has been shown1,10–24 that the energetically lowest

conical intersections responsible for fast deactivation to the

ground state are characterized by ring puckering modes.

The most favorable ring puckering conical intersections result

from the change of the HC6C5R (R = H, CH3) dihedral

angle in uracil and thymine and the HN1C2H dihedral

angle in adenine. This model also explains why by contrast

2-aminopurine, in which the C2 position is substituted by the

amino group, has a long excited state lifetime and strong

fluorescence.25–27 Alternatively, the puckering at the C6 atom

was also suggested to be responsible for adenine relaxation.28

Recently, surface-hopping dynamics studies of 4-amino-

pyrimidine (4-APy)29–31 and 9H-adenine24 showed that while

the former relaxes into the ground state via different conical

aDepartment of Chemistry, Stanford University, Stanford,CA 94305-5080, USA. E-mail: [email protected],[email protected], [email protected];Fax: (+1) 805-893-4120; Tel: (+1) 805-893-5921

b Institute of Chemistry, Eotvos Lorand University Budapest,1/A Pazmany P. stny., Hungary 1117

cDepartment of Chemistry and Biochemistry, University of California,Santa Barbara, CA-93106-9510, USA

dDepartment of Chemistry, University of the Pacific, Stockton,CA-95211, USA

eThe Institute of Organic Chemistry and Biochemistry,Flemingovo nam., 2, 166 10 Praha 6, Czech Republic

f Institute for Theoretical Chemistry, University of Vienna,Waehringerstrasse 17, A1090 Vienna, Austria

gMolecular and Biophysics Group, Institute for Molecules andMaterials, Radboud University, 6500 GL Nijmegen, The Netherlands

w Electronic supplementary information (ESI) available: The calcu-lated vertical ionization potentials for all the 2,4-diaminopyrimidinetautomers (Fig. S1 and Table S1). The vertical excitation energiescalculated at the RICC2 method with various basis sets (Table S2).The interpolation curves between the S1min_C2 and S1min_C6minima towards various MXS structures calculated at the CASSCFand MR-CISD(17)+Q methods (Fig. S2). This material is availablefree of charge on the Internet. See DOI: 10.1039/b917852j

This journal is �c the Owner Societies 2010 Phys. Chem. Chem. Phys., 2010, 12, 5375–5388 | 5375

PAPER www.rsc.org/pccp | Physical Chemistry Chemical Physics

Page 2: Effect of substituents on the excited-state dynamics of …devries/groupsite/pub/24DAPy 26DAPu.pdf · previously studied the nucleobase xanthine.46 In this paper we ... pseudo-Voigt

intersections formed by puckering at the C6, N1, and C2

atoms, the latter relaxes exclusively through conical inter-

sections formed by puckering at the C2 atom. This demon-

strates the role of hindering imposed by the imidazole ring of

adenine. Fig. 1 shows the structures and numbering of these

two compounds as well as the 2-amino substituted analogues

studied in the present paper, 2,4-diaminopyrimidine (2,4-DAPy)

and 2,6-diaminopurine (2,6-DAPu). Fig. 1b and c show the

lowest energy tautomers for the latter two compounds. One

may expect that a structural modification of the pyrimidine ring

in each of these molecules affects locations and accessibility of

conical intersections. In particular, amino-substitution in the C2

position should affect conical intersections associated with

puckering at that site, while addition of an imidazole ring to

form purine bases should affect conical intersections associated

with ring deformation at the C5 and C6 sites. Thus, in 2,4-DAPy

puckering at the C2 position is affected, and in 2,6-DAPu

puckering at both positions is modified. Therefore, it is especially

useful to compare the photochemistry of these compounds.

In addition to these theoretical aspects there exists a further

motivation for studying 2,4-DAPy and 2,6-DAPu bases. Joyce

et al. have proposed that alternative nucleobases, which may

form base pairs with geometries similar to Watson–Crick

structures, may have played a role in constructing the first

genetic code on the early Earth.32 Xanthine and 2,4-DAPy can

form an unnatural base pair that fits the Watson–Crick

geometry and can be incorporated into RNA and DNA by

polymerases.33–36 Another base pair mimicking the Watson–

Crick geometry is formed by 2,6-diaminopurine and uracil.

For each of these molecules, there exists a plausible, prebiotic

synthetic route.37–41

Purines and pyrimidines can exist in a variety of different

tautomeric forms, which can exhibit drastically different

photophysical behavior.42–45 Therefore, investigating the effects

of modification of the hetero-aromatic ring, in particular a

possible immobilization of certain parts of this ring by its

substitution at locations affecting the accessibility of conical

intersections can refine our understanding of the intrinsic

photostability of these alternative nucleobases.9,29–31

In the series of prebiotic and alternate RNA/DNA bases, we

previously studied the nucleobase xanthine.46 In this paper we

report the one-color R2PI spectra of 2,4-diaminopyrimidine

(2,4-DAPy) and 2,6-diaminopurine (2,6-DAPu) as well as

their IR-UV double resonance spectra. Because pyrimidine

derivatives have relatively high ionization energies unavailable

by two-photon ionization,47 we measured the ionization energy

of 2,4-DAPy by photoelectron spectroscopy to rule out three-

photon resonant ionization schemes. To explain the nature

and dynamics of the excited state of 2,4-DAPy, we charac-

terized the stationary points in its ground and excited states

and its S1/S0 conical intersections (minima on the crossing

seam, MXS) as well as pathways to internal conversion by

means of CASSCF, MR-CISD, and CASPT2 methods.

Experimental and theoretical methods

2,4-Diaminopyrimidine and 2,6-diaminopurine were purchased

from Sigma-Aldrich and used without further purification.

The experimental setups have been described in detail else-

where and only brief descriptions will be given here.48,49

Laser spectroscopy

The desorption laser, a Nd:YAG operating at 1064 nm, is

attenuated to 1 mJ cm�2 and focused to a spot approximately

0.5 mm diameter within 2 mm in front of the nozzle orifice. We

translate the sample in order to expose fresh sample to

successive laser shots. The nozzle consists of a pulsed valve

with a nozzle diameter of 1 mm and a backing pressure of

6 atm of argon drive gas. The neutral molecules are skimmed

and then ionized with a frequency doubled dye laser. We

detect the ions in a reflectron time-of-flight mass spectrometer.

We obtain resonant two-photon ionization (R2PI) spectra

by monitoring mass selected peaks while tuning the one-color,

two-photon ionization wavelength. We measure UV-UV

double resonance spectra with two laser pulses separated in

time by 200 ns. Ionization laser intensities are approximately

3 mJ pulse�1 and are strongly attenuated to avoid saturation.

The first pulse serves as a ‘‘burn’’ pulse, which removes the

ground state population and causes depletion in the ion signal

of the second ‘‘probe’’ pulse, provided both lasers are tuned to

a resonance of the same tautomer. IR-UV double resonance

spectra are obtained in an analogous way with the burn laser

operating in the near-IR region. IR frequencies are produced

in an OPO setup (LaserVision) pumped by a Nd:YAG

laser operating at its fundamental frequency. For this work,

we operated within the range of 3200–3800 cm�1, which

encompasses NH and OH modes. Typical IR intensities in the

burn region are 12 mJ pulse�1 and the bandwidth is 3 cm�1.

Excited state lifetimes are measured in a two-color, two-photon

ionization experiment. The pump and probe photons are

separated by a time delay of 100 ps to 100 ns, and the ion

signal is recorded as a function of the time delay. The lifetime of

the excited state is given as the time constant of the exponential

decay of the ion signal.

Photoelectron spectroscopy

He-I photoelectron spectra were recorded on a custom-built

ATOMKI ESA-32 instrument.49 The spectrometer is equipped

with a Leybold-Heraeus UVS 10/35 high-intensity helium

discharge photon source. The sample was introduced into

the ionization chamber via solid inlet probe at an elevated

temperature of 260 1C. The spectra were calibrated against Ar2P3/2 and 2P1/2 peaks. The maximum error in the ionization

energies is estimated to be less than 0.04 eV, and the energy

resolution of the He–I spectra was better than 40 meV (FWHM)

as determined from Ar 3p ionizations. In the extraction of

the vertical ionization potentials, Shirley background and

pseudo-Voigt peaks were fitted on the experimental spectra.

The adiabatic ionization potential can be estimated as the

onset of the ionization and has a somewhat higher uncertainty

than the vertical ionization potentials.

Quantum chemical calculations

Identifying the tautomers. To assign the IR-UV double

resonance spectra, we computed the structures of the possible

tautomers using density functional theory (DFT) and ab initio

5376 | Phys. Chem. Chem. Phys., 2010, 12, 5375–5388 This journal is �c the Owner Societies 2010

Page 3: Effect of substituents on the excited-state dynamics of …devries/groupsite/pub/24DAPy 26DAPu.pdf · previously studied the nucleobase xanthine.46 In this paper we ... pseudo-Voigt

Fig. 1 (a) Structures of the four related compounds discussed in the text. Calculations predict conical intersections for 4-Apyr due to ring

deformations at positions C2 and C6, C5, and N1. Amino substitution in the C2 position affects the conical intersection associated with that

position, while addition of the 5 membered ring for the purines affects the conical intersections associated with ring deformation in the other

positions. (b) The tautomers and numbering scheme of 2,4-diaminopyrimidine. Relative energies obtained at the (a) B3LYP/6-311+G(2d,p)

(b) MP2(FC)/6-311+G(2d,p) and (c) G3 levels are in kcal mol�1. (c) The tautomers and numbering scheme of 2,6-diaminopurine. Relative

energies obtained at the (a) B3LYP/6-311+G(2d,p) (b) MP2(FC)/6-311+G(2d,p) and (c) G3 levels are in kcal mol�1.

This journal is �c the Owner Societies 2010 Phys. Chem. Chem. Phys., 2010, 12, 5375–5388 | 5377

Page 4: Effect of substituents on the excited-state dynamics of …devries/groupsite/pub/24DAPy 26DAPu.pdf · previously studied the nucleobase xanthine.46 In this paper we ... pseudo-Voigt

methods. We combined Becke’s three-parameter hybrid

functional50 with the Lee–Yang–Parr exchange correlation

functional51,52 (B3LYP) and the 6-311+G(2d,p) basis set.53,54

We used the same basis set in frozen core second order

Møller–Plesset perturbation (MP2) calculations.55–60 We verified

the equilibrium structures by the absence of imaginary vibra-

tional frequencies. The calculated vibrational frequencies were

scaled by 0.961861 (DFT) or 0.949662 (MP2) to account for

electronic structure method deficiency and anharmonicity. To

establish the relative energies of the tautomers, we also applied

the Gaussian-3 (G3) composite method.63

Ionization energies. To assign the photoelectron spectrum,

we calculated ionization energies for 2,4-diaminopyrimidine,

using the B3LYP functional and 6-311+G(2d,p) basis set. We

calculated the adiabatic ionization potential as the energy

difference between the equilibrium geometries of the ionic

state and the ground state neutral molecule, estimating

the first vertical ionization energy as the difference between

the energies of the ionic state and the neutral molecule at the

equilibrium geometry of the latter. Then, at the equilibrium

geometry of the ion, we performed time-dependent DFT

(TD-DFT) calculations to compute the vertical excitation

energies of the ion.64–66 Vertical ionization energies are then

the sum of the adiabatic ionization energies and vertical excita-

tion energies of the ion.We also performed outer valence Green’s

function (OVGF) calculations with the 6-311+G(2d,p) basis set

at the equilibrium geometries of the neutral tautomers obtained

in the DFT calculations.67–74 All calculations were carried out

using the Gaussian 03 Rev. C. 02. quantum code package.75

Excited state investigations of 2,4-diaminopyrimidine. We

performed the calculations using the complete active space

self-consistent field (CASCCF) and multi-reference interaction

(MR-CI) methods for 2,4-DAPy. In the MR-CI approach,

single and double excitations from the CI reference space are

included (MR-CISD) and generalized interacting space restric-

tions are adopted.76 We constructed the orbital space for

CASSCF wavefunctions using 14 electrons in 10 orbitals, i.e.

composed of eight p orbitals and two lone pairs located on the

ring nitrogen atoms. We used a state-averaging procedure

using three states (SA-3) at the CASSCF level throughout

the calculations. Based on the CAS (14,10), we constructed the

MR-CISD reference space by moving orbitals with natural

occupation larger than 0.9 and smaller than 0.1 to the doubly

occupied and virtual spaces, respectively, resulting in a MR-CI

reference space composed of six electrons in five orbitals

(MR-CISD(6,5)). All single and double excitations from this

reference space were allowed; all core orbitals were frozen. To

reduce the computational cost of the calculations, up to a total

number of 17 doubly occupied orbitals were frozen in MRCI

calculations of the reaction paths. We tested this procedure

against full calculations for selected typical examples, taking

into account the size-consistency effects by means of Pople’s

correction method, indicated by +Q.77 We used the 6-31G**

basis set throughout the calculations.78,79 For comparison and

to verify the reliability of the applied basis set we also applied

the complete active space self-consistent-field second-order

perturbation theory (CASPT2) method80,81 with the same

reference CAS space and resolution-of-identity coupled cluster

to the second-order (RICC2)82,83 method.

We determined the minima on the ground and excited S1surfaces and on the seam of conical intersections (MXS) by

the CASSCF(14,10) method. We constructed reaction paths

between structures of the S1 minima and the MXSs using the

method of linear interpolation of internal coordinates (LIIC).

The CASSCF andMR-CISD energies of relevant points of the

reaction paths were plotted as a function of mass-weighted

distances between each point of the path and the S1 minimum.

We also determined the structures of the transition states for

selected reaction paths using the CASSCF(14,10) method. We

confirmed the character of the stationary point by Hessian

calculations within a selected space of internal coordinates

relevant for the ring puckering modes.

We simulated the absorption spectrum of 2,4-DAPy

employing the RICC2 and CASSCFmethods, using the Gaussian

broadening method described by Barbatti et al.84 We used

the same active space and number of states in the average

procedure as specified above. We generated five hundreds

points by a Wigner distribution in the ground vibrational state

of the ground electronic state, computed based on a Wigner

distribution for the ground vibrational and electronic state,

taking each nuclear degree of freedom within the harmonic

approximation. The Wigner distribution composed of 500

different geometries was projected onto the excited state by

multiplying it by the Einstein coefficient B computed for each

point. We used a phenomenological broadening of 0.05 eV.

We performed the optimization of conical intersections

using the analytic gradient and non-adiabatic coupling vectors85–89

available in the COLUMBUS program system.90–92 We per-

formed the CASPT2 calculations with the MOLCAS program

package93–95 and RICC2 computations were performed using

the Turbomole program system.96 The absorption spectrum

was simulated using the NEWTON-X program package.84,97

Results and discussion

2,4-Diaminopyrimidine R2PI and IR-UV double resonance

spectroscopy

Fig. 2 shows the R2PI spectrum of 2,4-diaminopyrimidine in

the wavelength range of 34 350-35 100 cm�1. In a one-color

experiment, we might not be able to observe the S1 ’ S0 origin

transition if the ionization potential of the molecule is too

high. Therefore, we determined the ionization potential by

photoelectron spectroscopy, as detailed below. For 2,4-diamino-

pyrimidine we found a relatively low vertical ionization

potential of 8.30 � 0.05 eV and an adiabatic ionization

potential of 7.86 � 0.05 eV, which corresponds to 33 471 cm�1

and 31 698 cm�1 photon wave numbers, respectively, in terms

of the one-color experiments. We, therefore, tentatively assign

the red-most peak at 34 459 cm�1 as the S1 ’ S0 origin

transition. If this assignment is correct, the intensity distri-

bution in the spectrum suggests significant geometric deforma-

tion of the excited state relative to the ground state.

The R2PI spectrum consists of a single tautomer, based on

IR-UV double resonance experiments at different UV probe

wavelengths, as detailed below. Fig. 3 shows the IR spectrum,

5378 | Phys. Chem. Chem. Phys., 2010, 12, 5375–5388 This journal is �c the Owner Societies 2010

Page 5: Effect of substituents on the excited-state dynamics of …devries/groupsite/pub/24DAPy 26DAPu.pdf · previously studied the nucleobase xanthine.46 In this paper we ... pseudo-Voigt

obtained at a probe frequency of 34 516 cm�1. The spec-

trum shows four distinct peaks in the frequency range of

3400–3600 cm�1. By fitting the experimental data with a sum

of Lorentzian curves, we extracted vibrational frequencies of

3452, 3464, 3563, and 3578 cm�1. This indicates the presence

of four free N–H stretching modes in the molecule and

suggests that we observe the diamino form (see Fig. 1b). We

computed the relative energies of all tautomers at the B3LYP/

6-311+G(2d,p), MP2(FC)/6-311+G(2d,p) and G3 levels. At

every level the di-amino form was the lowest energy tautomer

with the second most stable tautomer at a relative energy

15 kcal/mol higher than the diamino form. We calculated the

vibrational frequencies for all tautomers at the DFT and MP2

levels. The di-amino tautomer provides an excellent match

between the calculated and the experimental frequencies.

Table 1 summarizes the vibrational frequencies.

To determine that only the diamino tautomer is present in

the R2PI spectrum, we set the IR frequency to 3578 cm�1 and

scanned the R2PI spectrum again. The IR band at 3578 cm�1

is the most diagnostic band for the diamino tautomer

compared with the theoretically calculated IR bands for the

other tautomers. The R2PI spectrum obtained while hole

burning at this IR frequency showed a complete depletion in

the ion signal for the entire spectral range when the IR was on.

This result implies that all peaks in the R2PI spectrum are due

to the diamino tautomer.

Photoelectron spectroscopy

Fig. 4 shows the photoelectron spectrum of 2,4-diamino-

pyrimidine. The adiabatic ionization potential, derived from

the onset, is 7.86 � 0.05 eV. We derived the vertical ionization

potentials by fitting the sum of pseudo-Voigt curves on the

observed bands. Based on the intensities and the widths of the

bands, we find vertical ionization potentials corresponding to

six ionic states. An additional experimentally unresolved

broad band starts at an onset of 13 eV. It is implausible that

precise ionization energies can be extracted from the spectrum

above 13 eV, and we only note that the first maximum of this

range is at 13.51 eV.

Fig. 2 The R2PI spectrum of 2,4-diaminopyrimidine. The S1 ’ S0(pp*) origin is at 34 459 cm�1. Data were not corrected for variations in laser

intensities.

Fig. 3 The IR-UV double resonance spectra of 2,4-diaminopyrimidine and 2,6-diaminopurine in the N–H stretch IR wavelength range. The stick

spectra represent the best matching calculated vibrational frequencies at the B3LYP/6-311+G(2d,p) level. A scaling factor of 0.9618 was applied.

This journal is �c the Owner Societies 2010 Phys. Chem. Chem. Phys., 2010, 12, 5375–5388 | 5379

Page 6: Effect of substituents on the excited-state dynamics of …devries/groupsite/pub/24DAPy 26DAPu.pdf · previously studied the nucleobase xanthine.46 In this paper we ... pseudo-Voigt

Considering the relative energies of the tautomers of 2,4-

diaminopyrimidine, it is very unlikely that we observe more

than one tautomer in the gas phase, even at the applied inlet

temperature of 260 1C. However, we performed restricted

OVGF calculations at the 6-311+G(2d,p) level for all the

possible isomers, employing single point calculations on the

DFT equilibrium geometries. The outer valence Green’s func-

tion method is a useful tool in the assignment of photoelectron

spectra of organic molecules. The vertical ionization potentials

are usually predicted within 0.3 eV.98 We obtained the best

match with the experimental values for the most stable

tautomer, the diamino form. Fig. 4 lists the calculated ioni-

zation potentials. Photoelectron spectroscopy can also be

regarded as a tool for probing the ionic states. To obtain

adiabatic and vertical ionization energies for the diamino

form, we carried out TD-B3LYP/6-311+G(2d,p) calculations

at the equilibrium geometry of the ion. According to both

methods, the first peak of the spectrum can be assigned to the

ionization of the HOMO, that is, the removal of an electron

from the p system.

Table 2 lists the adiabatic and the lowest vertical ionization

energies. The predicted vertical ionization energies of all the

possible tautomers are compared to the experimental spectrum

in the Supporting Information (Fig. S1 and Table S1, ESI).w

2,6-Diaminopurine R2PI and IR-UV double resonance

spectroscopy

Previously, we reported a resonant two-photon ioniza-

tion spectrum of 2,6-diaminopurine between 32 000 cm�1

and 34 000 cm�1 without identifying the observed tautomer.

In the present study, we extended the R2PI spectrum up to

35 400 cm�1, as shown in Fig. 5. Two groups of peaks can be

distinguished, based on hole burning, with possible S1 ’ S0origins at 32 215 cm�1 and 34 881 cm�1, respectively. We

carried out IR-UV double resonance experiments on these

two possible origins. The resulting IR spectra, shown in Fig. 3,

are not identical, indicating the presence of at least two

different tautomers in the gas phase. Both spectra exhibit five

Table 1 Calculated vibrational frequencies for tautomers of 2,4-diaminopyrimidine in the N–H region. All frequencies have been scaled and givenin cm�1. MP2 scaling factor: 0.9496; B3LYP scaling factor: 0.9618. For numbering scheme, see Fig. 1. For comparison, the experimentalvibrational frequencies are 3452, 3464, 3563, and 3578 cm�1

Tautomer MP2 6-311+G(2d,p) o/cm�1 B3LYP 6-311+G(2d,p) o/cm�1 Vib. Mode

24DAP 3532 3574 N7H asdiamino tautomer 3518 3556 N8H as

3410 3456 N7H s3400 3443 N8H s

1H, 7 imino 3527 3564 N8H as3440 3483 N1H3403 3446 N8H s3351 3386 N7H

3H, 8 imino 3496 3534 N7H as3403 3445 N3H3387 3431 N7H s3346 3380 N8H

1H,3H 3467 3507 N1H7,8 imino 3422 3469 N3H

3359 3392 N10H3344 3378 N11H

3H, 7 imino 3535 N8H as3454 N3H3435 N8H s3387 N7H

1H, 8 imino 3508 N7H as3492 N1H3411 N7H s3328 N8H

5H, 8 imino 3593 N7H as3465 N7H s3316 N8H2986 CH

5H, 7 imino 3559 N8H as3436 N8H s3345 N7H2972 CH

Fig. 4 He–I photoelectron spectrum of 2,4-diaminopyrimidine.

pseudo-Voigt shaped peaks were fitted on the observed bands to extract

the vertical ionization potentials. The adiabatic ionization potential

was taken as the ionization onset at 7.86 eV. The adiabatic and

vertical ionization energies were also computed at the (TD)-B3LYP/

6-311+G(2d,p) and ROVFG/6-311+G(2d,p) levels and represented by

the stick spectra. See text for details.

5380 | Phys. Chem. Chem. Phys., 2010, 12, 5375–5388 This journal is �c the Owner Societies 2010

Page 7: Effect of substituents on the excited-state dynamics of …devries/groupsite/pub/24DAPy 26DAPu.pdf · previously studied the nucleobase xanthine.46 In this paper we ... pseudo-Voigt

IR bands. With the UV probe set at 34 881 cm�1 IR frequen-

cies appear at 3450, 3460, 3509, 3566, and 3576 cm�1. These

frequencies match very well with the calculated IR frequencies

of 2,6-diaminopurine.We assign the peaks at 3566 and 3576 cm�1

to the antisymmetric combinations of the N–H stretches in the

amino groups, the peaks at 3450 and 3460 cm�1 to the

symmetric combinations of the same stretches, and the peak

at 3509 cm�1 to either the N9–H or the N7–H stretching

mode. A hydrogen atom on the N9 position seems to be more

plausible, because it might have a smaller effect on the N–H

stretching modes of the amino groups, as discussed in the next

paragraph.

With the UV probe at 32 215 cm�1, the five IR bands appear

at 3428, 3460, 3513, 3533, and 3575 cm�1. When compared to

the former IR spectrum, significant red shifts of two peaks are

apparent. The peak at 3450 cm�1 is shifted to 3428 cm�1, and

the peak at 3566 cm�1 is shifted to 3533 cm�1. If the assign-

ment of the first IR spectrum is correct, this shift can be

explained by placing the hydrogen atom for this tautomer on

the N7 position instead of the N9 position. That would

perturb the N–H stretching modes in the N11 amino group.

DFT (B3LYP/6-311+G(2d,p)) and ab initio (MP2(FC)/

6-311+G(2d,p)) calculations of the vibrational frequencies

of the various tautomers support this qualitative assignment.

Table 3 shows the comparison of the calculated vibrational

frequencies with the experimental values. The two diamino

forms provide the best match. Furthermore, when calculating

the relative energies of the tautomers the N9H di-amino

tautomer is the most stable form, followed by the N7H

di-amino tautomer, at every level of calculation.

Excited state lifetimes

To obtain a measure of excited state lifetimes we performed

two color pump–probe experiments, in which we resonantly

excited the molecule to the S1 state, followed by ionization out

of the excited state with a 266 nm photon from a second laser,

with a variable delay between the two pulses. Fig. 6 shows

the results for 2,6-DAPu. Fitting the decay curves with a

single exponential decay, we obtained excited state lifetimes

of 6.3 � 0.4 ns for the N9H tautomer and 8.7 � 0.8 ns for the

N7H tautomer. Using laser pulses of about 5 ns pulse width

imposes a lower limit of the order of a few nanoseconds on

these lifetime measurements.

In the case of 2,4-DAPy, the lifetime is shorter than this

experimental limit so we were unable to perform two-color

experiments and our attempts to carry out UV-UV double

resonance experiments also failed. In this case we can use the

peak width in the R2PI spectrum to obtain a rough estimate of

the lower limit of the excited state lifetime. Fig. 7 shows the

peak at 34720 cm�1 together with simulations of the rotational

envelope with different Lorentzian linewidths. The simulated

lineshapes assume a dominantly b-type transition and a rota-

tional temperature of 20 K. These parameters are consistent

with our simulations of the peaks in the 2,6-DAPur spectrum,

for which the width is constrained by the experimental resolu-

tion only and not by lifetime broadening. The results, as

shown in Fig. 7, suggest an upper limit to the linewidth of

0.5 cm�1 corresponding to a lower limit of the excited state

lifetime of 10 ps. Therefore, we can bracket the experimental

window of our lifetime measurements for 2,4-DAPy roughly

between 10�11 and 10�9 s. This is clearly shorter than that of

2,6-DAPur, but it is longer than the lifetimes of 1.8 or 2.4 ps

and 5.2 or 6.4 ps, reported for uracil and thymine, respectively,

with excitation at 267 nm.99 We also note that the UV spectra

of uracil and thymine are broad while that of 2,4-DAPy is

sharp.5

This lifetime is clearly shorter than that of 2,6-DAPu, but

consistent with that of other pyrimidine bases excited at the

band origin.100,101 Not surprisingly, this 10 ps–1 ns time range

spans values much longer than the typical 1–6 ps found for

pyrimidine bases excited close to the first band maximum, well

above any barriers towards conical intersections.9,99 The sub-

nanosecond lifetime of band-origin excited 2,4-DAPy is a

strong indication that this molecule deactivates by means of

internal conversion just like 4-APy does. In order to examine

how the internal conversion takes place, we carried out

extensive analysis of the excited-state surfaces of this molecule,

which will be discussed in the next section.

Excited-state analysis of 2,4-diaminopyrimidine

Stationary points in the ground and excited states. Table 4

shows the vertical excitation energies calculated at the

CASSCF, MR-CISD, and CASPT2 levels. At the CASSCF

level, the first excited state is of pp* character while the secondexcited state is due to excitation from the lone pairs located on

Table 2 Experimental excitation and ionization energies (in eV) of2,4-diaminopyrimidine and 2,6-diaminopurine

2,4-diaminopyrimidine

Transition hv/eVS1 ’ S0 4.27D0 ’ S0 IEad 7.86

IEvert 8.30

7H-2,6-diaminopurine

S1 ’ S0 3.99

9H-2,6-diaminopurine

S1 ’ S0 4.32

Fig. 5 The R2PI spectra of the 7H and 9H tautomers of 2,6-

diaminopurine. The S1 ’ S0(pp*) origins are at 32 215 cm�1 and

34 881 cm�1, respectively. Asterisks indicate UV wavelengths that were

checked for ion signal depletion at diagnostic IR bands.

This journal is �c the Owner Societies 2010 Phys. Chem. Chem. Phys., 2010, 12, 5375–5388 | 5381

Page 8: Effect of substituents on the excited-state dynamics of …devries/groupsite/pub/24DAPy 26DAPu.pdf · previously studied the nucleobase xanthine.46 In this paper we ... pseudo-Voigt

the nitrogen atoms of the pyrimidine ring (np* transition).

The calculated energy gap between these states is 0.34 eV.

We obtained the same ordering of states at the CASPT2 and

RICC2 levels, with a smaller energy gap amounting to 0.2 and

0.1 eV, respectively. Increasing the flexibility of the basis set

does not change the ordering of the states and the energy

gap between them (see Table 4 and Supplementary Mate-

rial for CASPT2 and RICC2 results, respectively). At the

MR-CISD+Q level the ordering is reversed with an energy

gap of 0.3 eV.

The experimentally measured excitation energy at the band

origin is 4.27 eV. To estimate the energy of the band maxima

we have performed a simulation of the absorption spectra

(see Fig. 8 for the absorption spectrum simulated with RICC2

method). From the figure we calculated the energy difference

between the band maximum and band origin as 0.63 eV (the

same result was obtained employing the CASSCF method).

Adding this shift to the experimental value of the band origin

results in a band maximum at 4.91 eV, in a good agreement

with the calculated vertical excitation energies.

We found two minima on the S1 surface (Fig. 9a) using

CASSCF optimization, with energies of 4.75 and 5.08 eV

above the ground state minimum (Table 4). Both structures

show a distortion from the planarity of the ring which is more

pronounced in the latter case. For both minima, the CASSCF

wavefunctions show strong coupling between the np* and pp*configurations with almost equal weight. Unlike the case of

4-aminopyrimidine,29,30 we did not succeed in locating a third

minimum which is planar and of pp* character. Any attempt

to find this minimum ended up in conformation S1min_C6.

The energy ordering of S1min_C6 and S1min_C2 remains

the same at the MR-CISD(8)+Q level and at the CASPT2

level with an energy gap of about 0.2 and 0.3 eV, respectively.

The energy ordering is reversed when 17 orbitals are frozen in

the CI procedure, with the energetic changes being, however,

quite small. The latter calculations place the S1min_C2 below

the S1min_C6 by about 0.2 eV. The difference between the

results with 8 and 17 orbitals frozen is however not significant

for the discussion of reaction paths calculated by means of

linear interpolation curves and, thus, the less accurate but

Table 3 Calculated vibrational frequencies for tautomers of 2,6-diaminopurine in the N–H region. All frequencies have been scaled and givenin cm�1. MP2 scaling factor: 0.9496; B3LYP scaling factor: 0.9618. For numbering scheme see Fig. 1. For comparison, the experimentalvibrational frequencies are 3450, 3460, 3509, 3566, and 3576 cm�1 (UV probe 34 881 cm�1), and 3428, 3460, 3513, 3533, and 3575 cm�1 (UV probe32 215 cm�1)

Tautomer MP2 6-311+G(2d,p) o/cm�1 B3LYP 6-311+G(2d,p) o/cm�1 Vib. Mode

9H-26DAP 3505 3577 N11H asdiamino tautomer 3498 3567 N10H as

3438 3512 N9H3380 3456 N11H s3379 3452 N10H s

7H-26DAP 3493 3564 N10H asdiamino tautomer 3454 3519 N11H as + N7H

3356 3515 N7H + N11H s3375 3449 N10H s3346 3417 N11H s

1,9H, 11 imino 3525 N11H as3507 N9H3453 N1H3426 N10H s3377 N11H

1,7H, 11 imino 3508 N7H3500 N10H as3465 N1H3406 N10H s3351 N11H

3,9H, 10 imino 3586 N11H as3512 N9H3486 N3H3460 N11H s3391 N10H

1,9H, 10 imino 3543 N11H as3511 N9H3460 N1H3437 N11H s3392 N10H

3,7H, 10 imino 3522 N11H as + N7H3520 N7H + N11H as3484 N3H3417 N11H s3390 N10H

1,7H, 10 imino 3516 N11H as3485 N7H3460 N1H3422 N11H s3392 N10H

5382 | Phys. Chem. Chem. Phys., 2010, 12, 5375–5388 This journal is �c the Owner Societies 2010

Page 9: Effect of substituents on the excited-state dynamics of …devries/groupsite/pub/24DAPy 26DAPu.pdf · previously studied the nucleobase xanthine.46 In this paper we ... pseudo-Voigt

more computationally feasible approach with 17 orbitals

frozen can be used.

There is a substantial energetic relaxation of the first excited

states when moving from the Franck–Condon region to the

S1min_C6 and S1min_C2 optimized structures. This result is

consistent with the experimentally observed low intensity of

the 0–0 transition. Although the amount of relaxation energy

from 21A into the S1min_C2 structure, calculated with different

methods, is not as uniform as in the case of relaxation into the

S1min_C6 structure (see Table 4), results for all methods

used show that both minima located on the S1 surface are

energetically accessible from the Franck–Condon region.

Minima on the crossing seam. Table 4 lists the CASSCF and

MR-CISD energies and the character of five different minima

located on the S0/S1 crossing seam optimized at the CASSCF

level; Fig. 9b shows the corresponding structures. All struc-

tures display strong out-of-plane deformations with the twist

around either the CN or CC bonds. We adopted the Cremer-

Pople classification scheme in performing the conformational

analysis of these structures.102,103 Using this approach, we

identified four types of structure deformations, in particular

screw-boat (S), boat (B), half-chair (H) and envelope (E) con-

formations. The 3H4,1S2,

1S6 and B14 MXSs are energetically

grouped closely together at the CASSCF level within 0.4 eV;

all MXSs shown are energetically accessible starting from the

Franck–Condon region. The structures and the energetics of

all MXS points are very similar to those found by Barbatti

et al.29,30 for 4-aminopyrimidine although the results of

conformational analysis are not the same (see Table 4). This

situation indicates that there is only a small (if any) energetic

effect of the additional amino group on the puckering of

the pyrimidine ring and subsequent formation of conical

intersection.

In all cases there are two major configurations which

contribute to the CASSCF wavefunction (pp* and np*)(see Table 4). Although it is difficult to distinguish lone-

pair and p orbitals in such distorted structures, the main

character of the singly occupied orbitals is indicated in the

Table 4. The second configuration contributing to the wave-

function in all cases is the closed-shell configuration (CS). In

agreement with the results found for 4-aminopyrimidine, the

character of the singly occupied orbital was identified as

p orbital.

We performed MR-CISD calculations with 8 and 17

orbitals frozen in the CI procedure. The MR-CI and CASPT2

procedures resulted in relatively large splittings, usually

around 0.8 eV, of the S0 and S1 surfaces at the MXS

points determined at the CASSCF level. We estimated the

energies and MXS structures at the MR-CISD level from an

extrapolation of the LIIC curves (these results are obtained

from the MR-CISD(17)+Q calculations, see below). Table 4

lists the CASSCF and these estimated MR-CISD(17)+Q

results. When the dynamic correlation is included, there is a

change in the ordering of various MXS structures and the 1S2structure becomes the most stable one. However, all struc-

tures are placed in a relatively small energy range, with

the difference between the most and least stable structure

0.8 eV, in good agreement with the results of 4-aminopyrimidine.

Fig. 6 Exponential decay of the ion signal of 2,6-diaminopurine in

the two-color R2PI experiment. The extracted lifetime of the excited

state is 8.7 � 0.8 ns for the 7H tautomer and 6.3 � 0.3 ns for the 9H

tautomer.

Fig. 7 Detail of the 34 720 cm�1 peak in the 2,4-DAPy REMPI spectrum, compared with simulations of the rotational envelope at three different

Lorentzian linewidths.

This journal is �c the Owner Societies 2010 Phys. Chem. Chem. Phys., 2010, 12, 5375–5388 | 5383

Page 10: Effect of substituents on the excited-state dynamics of …devries/groupsite/pub/24DAPy 26DAPu.pdf · previously studied the nucleobase xanthine.46 In this paper we ... pseudo-Voigt

The CASPT2 energies, also estimated from the LIIC curves,

agree reasonably well with the MR-CI results. Thus, the

inclusion of the dynamic correlation effects does not change

the parameters for prediction of the dynamics of the molecular

systems studied.

Interpolation curves. We calculated LIIC curves for the

reaction paths between the S1min_C2 and S1min_C6 minima

and the four lowest MXS structures at the CASSCF and

MR-CISD levels. The LIIC curve for the reaction path S1min_C6

to 1S6 was calculated using the CASPT2 method as well. (See

Fig. 10 for the reaction path between S1min_C6 and 1S6. Other

reaction paths are reported in the ESI).w To make the studies of

reaction paths computationally more feasible we performed

MR-CISD calculations with 17 frozen orbitals. A comparison

of the character of the interpolation curves calculated at the

MR-CISD(8)+Q and MR-CISD(17)+Q levels for the reaction

path between S1min_C6 and 1S6 (see Fig. 10 and ESI)w shows

that the characters of the interpolation curves are very similar

justifying the use of the latter approach. The barriers for the

reaction paths estimated from the highest point of the inter-

polation curves are presented in Table 5. These values are

similar to those found by Zechmann and Barbatti for

4-APy.104 Since this estimation provides only an upper bound

for the reaction barrier, we have optimized the transition states

along the selected reaction paths, particularly the paths from

Table 4 Energies (in eV) of stationary points and conical intersections for 2,4-diaminopyrimidine relative to the ground state minimum structurecalculated at the CASSCF(14,10), MR-CISD(6,5)+Q and CASPT2 levels and comparison with 4-aminopyrimidine. The 6-31G** basis set wasused. The experimental excitation energy (band maximum) was estimated to be 4.91 eVa

CASSCF MR-CISD(17)+Qb CASPT2c 4APd MR-QDPT2

21A(pp*) DE 5.189 5.420 (5.406) 5.024 (4.938) 4.79 (5.34)fe 0.065 0.063

31A(np*) DE 5.528 5.044 (5.142) 5.204 (5.110) 4.71 (5.88)fe 0.013 0.012

S1min_C6 pp* + np* 4.754 4.743 (4.708) 4.573 (4.528) 4.22 (4.93)S1min_C2 pp* + np* 5.075 4.632 (4.917) 4.831 (4.822) 4.33 (5.02)

MXS(3H4) CS + pp* 4.594 4.2f 4.3 (4.4)f 4.62 (4.60)g

MXS (1S2) CS + pp* 4.645 3.9f 4.2 (4.2)f 4.36 (4.49)h

MXS (1S6) CS + pp* 4.759 4.5f 4.5 (4.5)f 4.71 (4.79)MXS (B1,4) CS + pp* 4.949 4.1f 4.3 (4.3)f (4.97)MXS(6E) CS + pp* 5.328 4.7f (5.42)

a For estimation of the experimental band maximum see text. b The results with eight frozen orbitals are given in parentheses. c The results

obtained with the 6-311G(2d,p) basis set are given in parentheses. d MR-QDPT2 energies (CASSCF energies are given in parentheses) calculated

for the relevant structures of 4-aminopyrimidine. e Oscillator strength. f Estimated from an extrapolation of the LIIC curves. g Assigned as 3S4 for

4-APy. h Assigned as E2 for 4-APy.

Fig. 8 Absorption spectrum of 2,4-DAPy calculated at the RICC2

method using the SVP basis set.

Fig. 9 (a) Structures of the two minima found on the S1 potential

energy surface optimized at the CASSCF(14,10)/6-31G** level. (b)

Structures of the five different minima located on the S0/S1 crossing

seam optimized at the CASSCF(14,10)/6-31G** level.

5384 | Phys. Chem. Chem. Phys., 2010, 12, 5375–5388 This journal is �c the Owner Societies 2010

Page 11: Effect of substituents on the excited-state dynamics of …devries/groupsite/pub/24DAPy 26DAPu.pdf · previously studied the nucleobase xanthine.46 In this paper we ... pseudo-Voigt

S1min_C6 towards 1S6 and3H4 MXS structures. The energies

of those structures are indicated in the graphs (see ESI)w and

in Table 5.

Relatively small barriers exist for the reaction paths between

the S1min_C6 and the 1S6 and B1,4 and3H4 MXSs. The only

sizeable barrier we found was for the reaction paths toward

the 1S2 structure. This finding reflects some immobilization of

the ring by substitution of the hydrogen atom by an amino

group at the C2 atom. The barriers connecting the S1min_C2

with MXSs are generally larger with values of approximately

0.8 eV. Starting from the highest points of the interpolation

curves we determined the true saddle points for the reaction

paths towards the 1S6 and 3H4 structures. At the CASSCF

and MR-CISD(17)+Q levels these calculations reduced the

reaction barriers to 0.17 and 0.12 eV at for 3H4, and to

0.04 and 0.02 eV for 1S6, respectively, indicating that both

conical intersections should be easily accessible leading to a

subsequent ultrafast deactivation to the ground state.

Comparing the results for 2,4-DAPy and those previously

reported for 4-APy shows that the additional amino-group

on the pyrimidine ring does not significantly influence the

energetics of the stationary points and conical intersections, as

well as the characters of interpolation curves. The only excep-

tion is the 1S2 structure and the interpolation curve towards it,

as a result of an immobilization of the pyrimidine ring at the

C2 position. Since 4-APy usually does not access the pathway

with C2 atom deformation, one would not expect significant

differences in the dynamics of these two systems, provided

comparable initial conditions are used. Therefore, in analogy

to the previously described surface hopping dynamics for

4-APy, the present reaction pathways for 2,4-DAPy indicate

that ultrafast internal conversion to the ground state should

be expected provided that sufficient initial kinetic energy is

available to surpass the small energy barriers to the 1S6 and3H4 conical intersections.

In the current experiments, the pump energy is in the range

of 34 000 to 35 000 cm�1 which corresponds to 4.2–4.3 eV.

Neglecting any vibrational corrections, this energy range is

close to or even below the calculated energies of the saddle

points (using the CASPT2 energies of MXSs and energy

barriers of the relevant saddle points collected in Table 4).

Therefore, the pump energy used in the experiment is not

sufficient to overcome the barrier to reach the conical inter-

sections, producing the much longer lifetime experimentally

observed for 2,4-DAPy.

When the relevant coordinates involve ring deformations,

lifetimes can be significantly different for specific structures. In

particular, if deformation involving C5 and C6 leads to

internal conversion in 2,4-DAPy then that pathway would

be significantly altered for 2,6-DAPu, in which the puckering

at C6 is strongly hindered. This observation explains why the

lifetime for 9H-adenine is longer than that calculated for

4-APy and also explains the very long lifetime (6–8 ns) for

2,6-DAPu, in which both the C2 position and the C5QC6

positions are modified. In this latter case the subtle difference

between the N7H and N9H tautomer lifetimes requires further

analysis. This effect of immobilization of the C5QC6 twist is

analogous to the findings of Zgierski et al., which showed that

Table 5 Barriers (in eV) located on the interpolation curves

CASSCF CISD(17)+Q

S1min_C6 to 1S6 0.145 Saddle point: 0.040 0.082 Saddle point: 0.024S1min_C6 to B1,4 0.419 0.153S1min_C6 to 3H4 0.407 Saddle point: 0.174 0.251 Saddle point: 0.118S1min_C6 to 1S2 1.041 1.042S1min_C2 to 1S2 0.614 0.976S1min_C2 to B1,4 0.686 0.974S1min_C2 to 1S6 0.788 0.830S1min_C2 to 3H4 0.870 0.857

Fig. 10 Reaction path between S1min_C6 and 1S6 calculated using

CASSCF(14,10), MR-CISD(8)+Q, and CASPT2 methods. Other

reaction paths are reported in the ESI.w

This journal is �c the Owner Societies 2010 Phys. Chem. Chem. Phys., 2010, 12, 5375–5388 | 5385

Page 12: Effect of substituents on the excited-state dynamics of …devries/groupsite/pub/24DAPy 26DAPu.pdf · previously studied the nucleobase xanthine.46 In this paper we ... pseudo-Voigt

in solution 5,6-trimethylenecytosine and 5,6-trimethyleneuracil

do not exhibit subpicosecond excited-state lifetimes charac-

teristic of the naturally occurring pyrimidine bases.21

Conclusions

We recorded resonant two-photon ionization spectra of 2,4-

diaminopyrimidine and 2,6-diaminopurine in the frequency

range of 32 000 to 36 000 cm�1. IR-UV double resonance

experiments supported by DFT and ab initio calculations

suggest that we observed the most stable tautomers of both

compounds. We also recorded and assigned the photoelectron

spectrum of 2,4-diaminopyrimidine, placing the adiabatic

ionization potential at 7.86 � 0.05 eV, and the first vertical

ionization potential at 8.30 � 0.04 eV. The S1 ’ S0 (pp*)origin of 2,4-diaminopyrimidine, as measured with one-color,

two-photon resonant ionization, is 34 459 cm�1. The excited

state lifetime measured at this pump wavelength is shorter

than the time resolution (approx. 1 ns) of the instrument and

longer than 10 ps, based on the sharp and resolved peaks in the

R2PI spectrum.

According to the IR-UV double resonance experiments,

both the N7H and N9H tautomers of 2,6-diaminopurine are

easily observable in the cold jet. The S1 ’ S0(pp*) origin is at

32 215 cm�1 for the N7H tautomer and 34 881 cm�1 for the

N9H tautomer. The measured excited state lifetime is 8.7� 0.8 ns

for the former and 6.3 � 0.4 ns for the latter.

This long lifetime is consistent with reduction of ring

deformations in both the C2 position and the C5QC6 twist,

affecting all conical intersections that lead to fast excited state

dynamics in 4-APy.

To understand the photodynamics of 2,4-diaminopyrimidine,

we carried out quantum chemical modeling. Based on the

results of the relative energies of stationary points on the S1surface and its crossings with the ground state at MXS points,

together with the nature of the interpolation curves we can

expect ultrafast dynamics in the excited state of 2,4-diamino-

pyrimidine which is predicted to be comparable to that of

4-aminopyrimidine. The additional NH2 group in the C2

position blocks the path to only one of the conical inter-

sections predicted for the analogous case of 4-aminopyrimidine,

namely the one associated with ring deformation at C2.

However, there are still other alternatives for almost barrier-

less access to conical intersections, associated with deformations

at the C5, C6, and N1 positions. The actually observed life-

times will depend on the chosen excitation energy which needs

to be large enough to surmount the existing small energy

barriers if ultrafast processes are to occur.

The present experiments show a significant reduction of the

lifetime of 2,4-DAPy as compared to the nanosecond lifetimes

determined for 2,6-DAPu, indicating a larger internal conver-

sion rate for the former molecule. However, the current

experiments involve a very narrow excitation window very

close to the band origin, and therefore exhibit the longest

possible lifetime. The pump energies in the range of 34 000 to

35 000 cm�1 are close to or possibly just below the computed

energy barriers, which explains the still rather large lifetime of

2,4-DAPy compared to that of other pyrimidine bases excited

at the center of the absorption band.

Acknowledgements

This study is based upon work supported by National Science

Foundation (CHE-0911564), and Hungarian National Science

Fund (OTKA T60669, F61153). Zsolt Gengeliczki gratefully

acknowledges the generous support of the Rosztoczy Founda-

tion. This work was supported by grants from the Ministry of

Education of the Czech Republic (Center for Biomolecules

and Complex Molecular Systems, LC512). It was part of

research project Z40550506 and by the Austrian Science Fund

within the framework of the Special Research Program F16

(Advanced Light Sources) and Project P18411-N19. Support

from the Preamium Academiae, Academy of Sciences of the

Czech Republic, awarded to PH in 2007, is gratefully

acknowledged.

References

1 C. M. Marian, J. Phys. Chem. A (USA), 2007, 111, 1545–1553.2 K. Seefeld, R. Brause, T. Haber and K. Kleinermanns, J. Phys.

Chem. A (USA), 2007, 111, 6217–6221.3 A. Abo-Riziq, L. Grace, E. Nir, M. Kabelac, P. Hobza and

M. S. de Vries, Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 20–23.4 A. L. Sobolewski, W. Domcke and C. Hattig, Proc. Natl. Acad.

Sci. U. S. A., 2005, 102, 17903–17906.5 B. B. Brady, L. A. Peteanu and D. H. Levy, Chem. Phys. Lett.,

1988, 147, 538–543.6 D. C. Luhrs, J. Viallon and I. Fischer, Phys. Chem. Chem. Phys.,

2001, 3, 1827–1831.7 H. Kang, B. Jung and S. K. Kim, J. Chem. Phys., 2003, 118,

6717–6719.8 C. Z. Bisgaard, H. Satzger, S. Ullrich and A. Stolow, Chem-

PhysChem, 2009, 10, 101–110.9 C. Canuel, M. Mons, F. Piuzzi, B. Tardivel, I. Dimicoli and

M. Elhanine, J. Chem. Phys., 2005, 122, 7.10 H. R. Hudock, B. G. Levine, A. L. Thompson, H. Satzger,

D. Townsend, N. Gador, S. Ullrich, A. Stolow andT. J. Martinez, J. Phys. Chem. A (USA), 2007, 111, 8500–8508.

11 T. Climent, R. Gonzalez-Luque, M. Merchan and L. Serrano-Andres, Chem. Phys. Lett., 2007, 441, 327–331.

12 L. Serrano-Andres, M. Merchan and A. C. Borin, Proc. Natl.Acad. Sci. U. S. A., 2006, 103, 8691–8696.

13 S. Perun, A. L. Sobolewski and W. Domcke, J. Am. Chem. Soc.,2005, 127, 6257–6265.

14 H. Chen and S. H. Li, J. Phys. Chem. A (USA), 2005, 109,8443–8446.

15 C. M. Marian, J. Chem. Phys., 2005, 122, 10.16 N. Ismail, L. Blancafort, M. Olivucci, B. Kohler andM. A. Robb,

J. Am. Chem. Soc., 2002, 124, 6818–6819.17 L. Blancafort, B. Cohen, P. M. Hare, B. Kohler and M. A. Robb,

J. Phys. Chem. A (USA), 2005, 109, 4431–4436.18 L. Blancafort and M. A. Robb, J. Phys. Chem. A (USA), 2004,

108, 10609–10614.19 L. Blancafort, J. Am. Chem. Soc., 2006, 128, 210–219.20 S. Matsika, J. Phys. Chem. A (USA), 2004, 108, 7584–7590.21 M. Z. Zgierski, S. Patchkovskii, T. Fujiwara and E. C. Lim,

J. Phys. Chem. A (USA), 2005, 109, 9384–9387.22 K. A. Kistler and S. Matsika, J. Chem. Phys., 2008, 128, 215102.23 K. A. Kistler and S. Matsika, J. Phys. Chem. A (USA), 2007, 111,

2650–2661.24 M. Barbatti and H. Lischka, J. Am. Chem. Soc., 2008, 130,

6831–6839.25 A. Broo, J. Phys. Chem., 1998, A102, 526–531.26 E. Nir, K. Kleinermanns, L. Grace and M. S. de Vries, J. Phys.

Chem. A (USA), 2001, 105, 5106–5110.27 S. Perun, A. L. Sobolewski and W. Domcke, Mol. Phys., 2006,

104, 1113–1121.28 E. Fabiano and W. Thiel, J. Phys. Chem. A (USA), 2008, 112,

6859–6863.29 M. Barbatti and H. Lischka, J. Phys. Chem. A (USA), 2007, 111,

2852–2858.

5386 | Phys. Chem. Chem. Phys., 2010, 12, 5375–5388 This journal is �c the Owner Societies 2010

Page 13: Effect of substituents on the excited-state dynamics of …devries/groupsite/pub/24DAPy 26DAPu.pdf · previously studied the nucleobase xanthine.46 In this paper we ... pseudo-Voigt

30 M. Barbatti, M. Ruckenbauer, J. J. Szymczak, A. J. A.Aquino and H. Lischka, Phys. Chem. Chem. Phys., 2008, 10,482–494.

31 M. Barbatti, B. Sellner, A. J. A. Aquino and A. Lischka,Nonadiabatic Excited-State Dynamics of Aromatic Heterocycles:Toward the Time-Resolved Simulation of Nucleobasis, SpringerScience, Business Media B.V., 2008.

32 G. F. Joyce, A. W. Schwartz, S. L. Miller and L. E. Orgel, Proc.Natl. Acad. Sci. U. S. A., 1987, 84, 4398–4402.

33 M. J. Lutz, J. Horlacher and S. A. Benner, Bioorg. Med. Chem.Lett., 1998, 8, 1149–1152.

34 M. J. Lutz, H. A. Held, M. Hottiger, U. Hubscher andS. A. Benner, Nucleic Acids Res., 1996, 24, 1308–1313.

35 J. A. Piccirilli, T. Krauch, S. E. Moroney and S. A. Benner,Nature, 1990, 343, 33–37.

36 S. Yamazaki, A. L. Sobolewski and W. Domcke, Phys. Chem.Chem. Phys., 2009, 11, 10165–10174.

37 H. J. Cleaves, K. E. Nelson and S. L. Miller, Naturwissenschaften,2006, 93, 228–231.

38 J. P. Ferris, O. S. Zamek, A. M. Altbuch and H. Freiman, J. Mol.Evol., 1974, 3, 301–309.

39 S. Miyakawa, H. J. Cleaves and S. L. Miller, Origins Life Evol.Biosphere, 2002, 32, 209–218.

40 M. P. Robertson, M. Levy and S. L. Miller, J. Mol. Evol., 1996,43, 543.

41 R. Saladino, C. Crestini, V. Neri, J. R. Brucato, L. Colangeli,F. Ciciriello, E. Di Mauro and G. Costanzo, ChemBioChem,2005, 6, 1368–1374.

42 E. Nir, M. Muller, L. I. Grace and M. S. de Vries, Chem. Phys.Lett., 2002, 355, 59–64.

43 E. Nir, C. Janzen, P. Imhof, K. Kleinermanns and M. S. de Vries,J. Chem. Phys., 2001, 115, 4604–4611.

44 E. Nir, I. Hunig, K. Kleinermanns and M. S. de Vries, Phys.Chem. Chem. Phys., 2003, 5, 4780–4785.

45 K. A. Seefeld, C. Plutzer, D. Lowenich, T. Haber, R. Linder,K. Kleinermanns, J. Tatchen and C. M. Marian, Phys. Chem.Chem. Phys., 2005, 7, 3021–3026.

46 M. P. Callahan, B. Crews, A. Abo-Riziq, L. Grace, M. S. deVries, Z. Gengeliczki, T. M. Holmes and G. A. Hill, Phys. Chem.Chem. Phys., 2007, 9, 4587–4591.

47 R. Tembreull, C. H. Sin, H. M. Pang and D. M. Lubman, Anal.Chem., 1985, 57, 2911–2917.

48 G. Meijer, M. S. de Vries, H. E. Hunziker and H. R. Wendt, Appl.Phys. B: Photophys. Laser Chem., 1990, 51, 395–403.

49 B. Csakvari, A. Nagy, L. Zanathy and L. Szepes, Magy. Kem.Foly., 1992, 98, 415–419.

50 A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.51 C. T. Lee, W. T. Yang and R. G. Parr, Phys. Rev. B: Condens.

Matter, 1988, 37, 785–789.52 B. Miehlich, A. Savin, H. Stoll and H. Preuss, Chem. Phys. Lett.,

1989, 157, 200–206.53 A. D. Mclean and G. S. Chandler, J. Chem. Phys., 1980, 72,

5639–5648.54 R. Krishnan, J. S. Binkley, R. Seeger and J. A. Pople, J. Chem.

Phys., 1980, 72, 650–654.55 C. Moller and M. S. Plesset, Phys. Rev., 1934, 46, 618–622.56 J. A. Pople, M. Headgordon and K. Raghavachari, J. Chem.

Phys., 1989, 90, 4635–4636.57 M. J. Frisch, M. Headgordon and J. A. Pople, Chem. Phys. Lett.,

1990, 166, 275–280.58 M. J. Frisch, M. Headgordon and J. A. Pople, Chem. Phys. Lett.,

1990, 166, 281–289.59 M. Headgordon and T. Headgordon, Chem. Phys. Lett., 1994,

220, 122–128.60 S. Saebo and J. Almlof, Chem. Phys. Lett., 1989, 154, 83–89.61 M. P. Andersson and P. Uvdal, J. Phys. Chem. A (USA), 2005,

109, 2937–2941.62 A. P. Scott and L. Radom, J. Phys. Chem., 1996, 100,

16502–16513.63 L. A. Curtiss, K. Raghavachari, P. C. Redfern, V. Rassolov and

J. A. Pople, J. Chem. Phys., 1998, 109, 7764–7776.64 R. E. Stratmann, G. E. Scuseria and M. J. Frisch, J. Chem. Phys.,

1998, 109, 8218–8224.65 R. Bauernschmitt and R. Ahlrichs, Chem. Phys. Lett., 1996, 256,

454–464.

66 M. E. Casida, C. Jamorski, K. C. Casida and D. R. Salahub,J. Chem. Phys., 1998, 108, 4439–4449.

67 J. V. Ortiz, J. Chem. Phys., 1988, 89, 6353–6356.68 L. S. Cederbaum, J. Phys. B: At. Mol. Phys., 1975, 8, 290–303.69 W. von Niessen, J. Schirmer and L. S. Cederbaum, Comput. Phys.

Rep., 1984, 1, 57.70 V. G. Zakrzewski and W. Vonniessen, J. Comput. Chem., 1993,

14, 13–18.71 V. G. Zakrzewski and J. V. Ortiz, Int. J. Quantum Chem., 1995,

53, 583–590.72 J. V. Ortiz, Int. J. Quantum Chem., 1988, 22, 431.73 J. V. ortiz, Int. J. Quant. Chem. Symp., 1989, 23, 321.74 J. V. Ortiz, V. G. zakrzewski and O. Dolgounircheva, in

Conceptual Perspectives in Quantum Chemistry, ed. J. L. Calaisand E. S. Kryachko, Kluwer Academic, 1997, p. 465.

75 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,M. A. Robb, J. R. Cheeseman, J. Montgomery, J. A., T. Vreven,K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi,V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota,R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox,H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo,J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev,A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski,P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador,J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich,A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick,A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz,Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov,G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin,D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng,A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson,W. Chen, M. W. Wong, C. Gonzalez and J. A. Pople, Gaussian,Inc., Wallingford, CT, 2004.

76 A. Bunge, J. Chem. Phys., 1970, 53, 20.77 J. A. Pople, R. Seeger and R. Krishnan, Int. J. Quantum Chem.,

1977, 149–163.78 W. J. Hehre, R. Ditchfie and J. A. Pople, J. Chem. Phys., 1972,

56, 2257.79 J. S. Binkley, J. A. Pople and W. J. Hehre, J. Am. Chem. Soc.,

1980, 102, 939–947.80 K. Andersson, P. A. Malmqvist and B. O. Roos, J. Chem. Phys.,

1992, 96, 1218–1226.81 K. Andersson, P. A. Malmqvist, B. O. Roos, A. J. Sadlej and

K. Wolinski, J. Phys. Chem., 1990, 94, 5483–5488.82 A. Kohn and C. Hattig, J. Chem. Phys., 2003, 119, 5021–5036.83 C. Hattig, J. Chem. Phys., 2003, 118, 7751–7761.84 M. Barbatti, G. Granucci, M. Persico, M. Ruckenbauer,

M. Vazdar, M. Eckert-Maksic and H. Lischka, J. Photochem.Photobiol., A, 2007, 190, 228–240.

85 M. Dallos, H. Lischka, R. Shepard, D. R. Yarkony andP. G. Szalay, J. Chem. Phys., 2004, 120, 7330–7339.

86 H. Lischka, M. Dallos, P. G. Szalay, D. R. Yarkony andR. Shepard, J. Chem. Phys., 2004, 120, 7322–7329.

87 H. Lischka, M. Dallos and R. Shepard, Mol. Phys., 2002, 100,1647–1658.

88 R. Shepard, H. Lischka, P. G. Szalay, T. Kovar andM. Ernzerhof, J. Chem. Phys., 1992, 96, 2085–2098.

89 R. Shepard, in Modern Electronic Structure Theory, ed.D. R. Yarkony, World Scientific, Singapore, 1995, vol. 1,p. 345.

90 H. Lischka, R. Shepard, R. M. Pitzer, I. Shavitt, M. Dallos,T. Muller, P. G. Szalay, M. Seth, G. S. Kedziora, S. Yabushitaand Z. Y. Zhang, Phys. Chem. Chem. Phys., 2001, 3, 664–673.

91 H. Lischka, R. Shepard, F. B. Brown and I. Shavitt, Int. J.Quantum Chem., 1981, 91–100.

92 H. Lischka, R. Shepard, I. Shavitt, R. M. Pitzer, M. Dallos,T. Muller, P. G. Szalay, F. B. Brown, R. Ahlrichs, H. J. Boehm,A. Chang, D. C. Comeau, R. Gdanitz, H. Dachsel, C. Ehrhardt,M. Ernzefhof, P. Hoechtl, S. Irle, G. Kedziora, T. Kovar,V. Parasuk, M. J. M. Pepper, P. Sharf, H. Shiffer,M. Schindler, M. Schuler, M. Seth, E. A. Stahlberg,J.-G. Zhao, S. Yabushita, Z. Zhang, M. Barbatti, S. Matsika,M. Schuurmann, D. R. Yarkony, S. R. Brozell, E. V. Beck and

This journal is �c the Owner Societies 2010 Phys. Chem. Chem. Phys., 2010, 12, 5375–5388 | 5387

Page 14: Effect of substituents on the excited-state dynamics of …devries/groupsite/pub/24DAPy 26DAPu.pdf · previously studied the nucleobase xanthine.46 In this paper we ... pseudo-Voigt

J.-P. Blaudeau, Columbus, an ab initio electronic structure programrelease 5.9.1, www.univie.ac.at/columbus, 2006.

93 G. Karlstrom, R. Lindh, P. A. Malmqvist, B. O. Roos, U. Ryde,V. Veryazov, P. O. Widmark, M. Cossi, B. Schimmelpfennig,P. Neogrady and L. Seijo, Comput. Mater. Sci., 2003, 28,222–239.

94 P. A. Malmqvist, A. Rendell and B. O. Roos, J. Phys. Chem.,1990, 94, 5477–5482.

95 B. O. Roos and P. R. Taylor, Chem. Phys., 1980, 48, 157–173.96 R. Ahlrichs, M. Bar, M. Haser, H. Horn and C. Kolmel, Chem.

Phys. Lett., 1989, 162, 165–169.97 M. Barbatti, G. Granucci, M. Ruckenbauer, J. Pittner, M. Persico

and H. Lischka, NEWTON-X: a package for Newtonian dynamicsclose to the crossing seam, www.univie.ac.at/newtonx, 2007.

98 I. P. Csonka, U. Szepes and A. Modelli, J. Mass Spectrom., 2004,39, 1456–1466.

99 H. Kang, K. T. Lee, B. Jung, Y. J. Ko and S. K. Kim, J. Am.Chem. Soc., 2002, 124, 12958–12959.

100 Y. G. He, C. Y. Wu and W. Kong, J. Phys. Chem. A (USA),2003, 107, 5145–5148.

101 Y. G. He, C. Y. Wu and W. Kong, J. Phys. Chem. A (USA),2004, 108, 943–949.

102 D. Cremer, Acta Crystallogr., Sect. B: Struct. Sci., 1984, 40,498–500.

103 D. Cremer and J. A. Pople, J. Am. Chem. Soc., 1975, 97,1354–1358.

104 G. Zechmann and M. Barbatti, Int. J. Quantum Chem., 2008, 108,1266–1276.

5388 | Phys. Chem. Chem. Phys., 2010, 12, 5375–5388 This journal is �c the Owner Societies 2010